Human spinal cord injury (SCI) results in permanent functional impairments. A loss in mobility is one of the most noticeable and debilitating consequences. Activity-based treadmill (TM) training attempts to promote recovery of walking by providing afferent sensory input to spinal central pattern generators (CPGs). This approach shows great potential, as activity-dependent plasticity is abundant in the spinal cord. With training, spinal neurons relearn to initiate components of locomotion despite lost descending drive. To induce the most robust learning and recovery after SCI, we hypothesize that training must occur early to take advantage of peak CNS plasticity. Unfortunately, training interventions that are delivered too early fail to produce functional improvement and even disrupt neurovascular integrity at the epicenter. Direct cellular impediments remain unclear, as the sequela of SCI is complex and leads to a host of dysfunction. Importantly, inflammatory mechanisms spread to progressively greater distances from the lesion site acutely after injury. It is clear that glial reactivity and cytokine production are central to secondary pathogenesis after SCI. The influence of these processes on motor relearning within regions of locomotor CPGs remains unexamined. Does SCI induce an early microenvironment that prevents motor relearning? A potent mediator of early pathology at the injury site is matrix metalloproteinase-9 (MMP-9). At the epicenter, MMP-9 facilitates leukocyte infiltration and cleaves a number of cytokines and chemokines. In this proposal, we present the first evidence of MMP-9 activity away from the injury in the lumbar cord during acute stages. We intend to test the hypothesis that contusive SCI results in early remote production of MMP-9 that prevents sparing-induced plasticity and motor re-learning. Our strategy involves manipulating remote activity of MMP-9 in conjunction with early treadmill training after SCI. Using a novel spinal learning paradigm; we will examine the isolated capacity of lumbosacral segments to determine plasticity of locomotor CPGs. Over the course of our experiments, we will identify an optimal environment for motor relearning. Our preliminary data suggests that removal of MMP-9 attenuates remote inflammation and in combination with early TM training promotes adaptive plasticity and robust locomotor recovery. Findings from our work will provide therapeutic potential for an isolated treatment to the lumbar cord injury in conjunction with early TM training.